Features - MurmurTone | Privacy-First Voice-to-Text

Privacy & Security

Your Voice Stays Where It Belongs

Built from the ground up with privacy as the core principle, not an afterthought.

Your Voice Never Leaves Your Machine

Every sound you speak is processed by AI running directly on your computer. There's no server to send your data to - because it never goes anywhere. The recognition engine runs locally, the text stays local, and your privacy is guaranteed by architecture, not promises.

Invisible to the Cloud

No transcripts stored on remote servers. No audio files uploaded for "quality improvement." No usage data collected for advertising. MurmurTone doesn't even have a way to know what you said - there's no telemetry, no analytics, no tracking of any kind.

Anywhere You Go

Airplanes at 35,000 feet. Secure government facilities. Remote mountain cabins. Offshore drilling platforms. MurmurTone works exactly the same with zero internet connection - because it was designed that way from day one. No network means no vulnerabilities.

Speed & Productivity

Faster Than the Cloud

No server round-trips means instant results. Speak and watch words appear.

Results in Under a Second

No server latency. No upload time. No waiting for cloud processing. Speak, release your hotkey, and words appear instantly in any application.

<1s average latency

AI Text Cleanup Coming Soon

Local AI will polish your transcription automatically: fixing grammar and adjusting formality, all without sending a single character to the cloud. Coming in a future update. (Filler-word removal is available today, see below.)

100% local AI

Types Into Any App

Word, Slack, your browser, email clients, code editors, games - MurmurTone types directly into whatever window is active. No copy-paste, no app switching, just speak and it's there.

Universal compatibility

Power Features

Built for Professionals

Advanced capabilities that adapt to how you work.

GPU Acceleration

Harness your graphics card for dramatically faster transcription. NVIDIA GPU owners see up to 10x speed improvements over CPU-only processing.

Voice Commands

Say "new line", "new paragraph", "period", or "select all" to control your text hands-free. Natural commands that feel like second nature.

17+ Languages

English, Spanish, French, German, Japanese, Chinese, Korean, and 10 more languages, plus auto-detect for the rest. All processed locally.

Translation Mode

Speak in any supported language, get English text output. Perfect for multilingual environments or non-native speakers drafting in English.

Multiple AI Models

Choose from tiny (fastest) to large (most accurate). Pick the right balance for your hardware and accuracy needs.

Custom Vocabulary

Teach MurmurTone your jargon. Medical terms, legal language, company names, technical acronyms - add custom words that get recognized correctly.

Why MurmurTone?

Feature	MurmurTone	Cloud Services
Data Privacy	✓ Nothing leaves your PC	✗ Uploaded to servers
Works Offline	✓ Always available	✗ Requires internet
Response Time	✓ Under 1 second	✗ 2-5+ seconds typical
Local AI Cleanup	Coming soon	✗ Cloud-only or none
Languages	✓ 17+ local	✓ Varies by service
Translation	✓ Local processing	✗ Cloud-dependent
Custom Vocabulary	✓ Included	✗ Paid tier or limited
Annual Cost	$49/year	$96-180/year

Customization

Make It Yours

Fine-tune every aspect to match your workflow.

Flexible Hotkeys

Any key combination you want. Ctrl+Space, F13, or a mouse button - your choice.

Filler Word Removal

Automatically strip "um", "uh", "like", and other verbal fillers from your transcripts.

Recording Modes

Push-to-talk, toggle on/off, or auto-stop when you pause. Pick what feels natural.

Audio File Transcription

Drop in audio files to transcribe meetings, interviews, or recordings. Same privacy, same accuracy.

Privacy Isn't a Feature.
It's the Foundation.